Shared-Parameter Closure of CSL Collapse Models from Public Upper-Bound Evidence: Compression Results and Evidence-Class Limits

This Zenodo record contains the main preprint and supplementary information for a shared-parameter closure analysis of Continuous Spontaneous Localization (CSL) using public upper-bound evidence and direct astrophysical inequalities. Over the reporting domain x = log10 (rC / m) in -10, 0 and y = log10 (λ / s^-1) in -20, -5, the study reconstructs admissible upper envelopes for XENONnT, a GW/LISA composite constraint, radiation-channel searches, cold-atom and bulk-heating constraints, cantilever-related bounds, and direct astrophysical heating inequalities. The principal result is a quantified compression of the shared admissible CSL region under the current public evidence stack, together with an evidence-class limit: when each channel is represented only by one-sided admissible upper envelopes or direct inequalities, the shared-envelope admissible set is exactly the intersection of the individual admissible sets. Under this evidence class alone, positive-measure closure-failure regions cannot be established. Accordingly, the record supports a bounded claim: current public single-channel summaries overstate the shared admissible CSL volume, whereas decisive inconsistency testing requires richer, likelihood-aware evidence. Files included in this record: 1. Main preprint PDF. 2. Supplementary PDF containing source/provenance notes, formulas, exact reported metrics, and reproducibility notes.